Résumé
This article studies the epidemiological and economic impacts of a universal testing and treatment policy of Human Immunodeficiency Virus (HIV) in South Africa. A model of disease transmission is built to simulate several implementation scenarios of the policy. Different behavioral responses in the general population are considered. The results show that the success of a large-scale HIV testing and treatment program in South Africa depends on its implementation conditions. The policy can lead to a reduction of the HIV epidemic, even in the case of a large relapse in preventive behaviors in the general population, if implementation conditions are favorable. This is the case if the number of infected individuals who are infectious is greatly reduced. From an economic point of view, taking into account the positive externality of antiretroviral (ARV) treatments changes the traditional framework of cost-benefit analyses. A large-scale testing and treatment program would be cost-saving in the case of favorable implementation conditions, even following a large increase in risk behaviors after the scaling up of ARV treatments. By contrast, the analysis stresses out the potential perverse effects of scaling up ARV treatments in South Africa if the intervention is set up without ensuring enough resources for patients' monitoring and the availability of effective ARV drugs. Indeed, if the number of treated patients rises while adherence of patients to treatments decreases and the rate of loss to follow-up increases, the policy could extend the pool of infectious patients and lead to a long-term amplification of the epidemic.